One of the major downsides of AI companions like ChatGPT or Gemini at the moment is the safety limitations you may find yourself encountering. You ask an AI a tricky question, and it responds with a polite, hyper-cautious disclaimer. It apologizes for its limitations, sticks firmly to its safety guidelines, and often leaves you feeling like you’re talking to a well-mannered but extremely boring diplomat.
In comes Grok. Created by Elon Musk’s AI company xAI, Grok is designed to be a bit more of a rebellious AI chatbot in a world of overly courteous chatbots. Not only does it answer questions but it does so with a fair amount of snark and personality.
Another bonus of the Grok companion is that it has full access to the X social media platform (formerly Twitter). It was originally only accessed via, and trained on, X, but now has its own application. Here’s everything you need to know about Grok AI.
What is Grok?
Grok is a generative AI chatbot developed by xAI, Musk’s artificial intelligence startup. It was originally built atop the Grok-1 model, which was developed over the course of months on a cluster of “tens of thousands” of GPUs and leveraged the Flux.1 model by Black Forest Labs for its image-generation capabilities.
The model was trained on a mix of web data and X user data and unlike other AI companions, it doesn’t have a traditional knowledge cut off point as it has real time access to the internet through X.
This chatbot is fast, often hilarious, and built to tackle the “spicy” questions that competing models refuse to touch. Whether you love or loathe Musk, Grok represents a major philosophical and technical departure in the AI landscape.
The chatbot first rolled out in November 2023 (and, surely coincidentally, less than a year after Musk signed an open letter demanding the industry take a six-month moratorium toward further AI development).
If you’re wondering where the bizarre name came from, “Grok” was coined in the Robert Heinlein novel Stranger in a Strange Land. The original meaning was to “understand intuitively or by empathy, to establish rapport with.”
Where did Grok come from?
Elon Musk co-founded OpenAI along with Sam Altman and other investors in 2015. However, just three years later, Musk quit the startup, claiming he “didn’t agree with some of what OpenAI team wanted to do.”
Altman and company went on to launch ChatGPT in November 2022, kicking off the AI revolution, and giving Musk a wicked case of FOMO.
In April 2023, just a month after OpenAI released GPT-4, Musk went on the Tucker Carlson Show to announce he would be building “a maximum truth-seeking AI that tries to understand the nature of the universe” that wouldn’t bow to social niceties like ChatGPT, which, he argued, was being “trained to be politically correct.”
He called the AI “TruthGPT” because, as has proven time and again, branding is not Musk’s strong suit. TruthGPT was renamed Grok at its release.
Grok-1 was superseded by Grok-1.5 in March 2024, which offered improved performance and a context length of 128,000 tokens. That April, X incorporated Grok AI into its Explore section to summarize breaking news stories, a role previously performed by humans.
On its first day in use, the AI immediately hallucinated a headline about Iran bombarding Israel with “heavy missiles” that was then promoted by the company’s trending news tab. Grok-1.5 was then replaced by Grok-2, and Grok-2 mini, in August 2024.
Grok-3 then released in February 2025 and was subsequently replaced by Grok-4 in July 2025, which is now the current model.
What can Grok do?
Grok has been “modeled after The Hitchhiker’s Guide to the Galaxy,” the company explained in its launch announcement, “so [it’s] intended to answer almost anything.” The open-source large language model reportedly boasts 314 billion parameters (about three-quarters the size of Llama 3.1-405B) and has been trained on both data from the public web as well as the data from X users themselves.
Grok can answer user queries on subjects up to its October 31 2024 knowledge cutoff. However, it can still answer questions surrounding events after that cutoff as Grok can perform web-searches as well as use “real-time access” to find information on X.
However, in turn that may be why Grok appears more susceptible to hallucinations and repeating misinformation than other AIs like ChatGPT, since a lot of it’s information is sourced from a social media platform.
Grok’s defining feature is its willingness to discuss subjects that are considered taboo by other chatbots and will readily discuss issues of politics, religion, and race with its users.
For example, when one X employee asked for a vulgar response to the question, “When is it appropriate to listen to Christmas music?” the AI responded with “whenever the hell you want” and added that those who disagree should “shove a candy cane up their ass and mind their own damn business.”
Similarly, Grok’s image generation capabilities, which debuted in August 2024 alongside Grok-2 and mini, have very few guardrails. Grok 2.0 claims to have guardrails when asked, but when releasing almost everything was fair game.
Users are able to generate images of celebrities, politicians, and other public figures as well as copyrighted assets like Disney’s Mickey Mouse and Nintendo’s Mario.
It should be noted that neither Disney nor Nintendo are renowned for their tolerance of IP theft and with Disney suing Midjourney in the past. However, Grok has been seemingly left alone for the time being. As of right now, no legal action has been taken and you can still generate images of anything you want, including copyright protected characters, but how you use these images might pose issues.
Overall, Grok can do pretty much anything that its physically capable of doing – unfortunately its lack of interface (such as actual working arms and legs) prevent it from completing physical chores however it can explain exactly what you need to do to complete these tasks.
It won’t be able to cook you a meal but if you ask it for step by step directions on how to do it yourself, it can provide you with these. You can use Grok to create poems, edit lines of code, write you emails and so much more.
One of the main features of Grok is its integration into X formerly known as Twitter, as it allows you to tag the AI beneath tweets and ask it to fact check, explain or clarify the contents of the tweet.
How much does Grok cost?
Grok can be accessed for free so long as you have an X account, however the free version is limited. You’re only able to input approximately 10 queries every 2 hours, however this isn’t a strict limit and can vary based on traffic.
You can unlock all of the features of Grok via multiple different subscription options. X premium, which comes in at $40 per month, includes unlimited access to Grok as well as a range of other X benefits including an ad-free experience and a revenue share program.
SuperGrok is another subscription service which comes in at $30 a month or $300 a year. This includes advanced features like DeepSearch as well as enhanced reasoning modes. For those who aren’t too fussed about the X platform but still want access to all of the features that Grok can offer, then this is likely the plan for you.
Grok 4 Heavy is the final tier which is available, coming in at $300 per month. This is the most powerful version of Grok 4.
After releasing Grok-3 back in February, X premium members saw subscription prices increase significantly, with subscriptions originally starting at $8 per month.
What’s the latest version of Grok?
Grok 4 is the latest version of the AI, which was released back at the start of July 2025. When releasing this new update, X stated that Grok 4 is the most intelligent model in the world.
You have the ability for native tool use and real-time search integration, features which have already been available within ChatGPT for a while now.
X also states that Grok 4 has scaled up reinforcement learning. This has refined Grok’s reasoning abilities making it more understanding and allows it to provide more human like responses.
Which devices can use Grok?
In January 2025, Grok made the leap from being a feature within the X chat platform to being an app of its own with the release of Grok for iOS. Available on both iOS and iPadOS devices, the app offers free access to Grok-2, which was released in August 2024.
Users can leverage the AI’s text and image generation capabilities through the app, though the existing limitations for free tier users (10 queries every two hours) remain in place. This means you can use Grok on iOS, Andriod and web within X.
On top of this, Grok has been implemented into Tesla vehicles as of July 12 2025, allowing you to talk to the AI hands free. You can choose Grok’s voice and personality to enhance your convince on the go.
What’s next for Grok?
A new Grok is seemingly released every 6 months which suggests that Grok 5 may be coming around the end of this year or early next year. We also know at the moment that Grok will become even more heavily integrated with X as it will steer your feed and also allow you to ask it to tune what you see. This will be rolling out around the end of the year.
Other than this, we don’t have very much information on what is coming next for Grok and with so many new features implemented recently, it makes sense that development may be winding down for a short while.
